Rationale
Applying
knowledge learned in one task or domain (the source) to improve
learning on another related task or domain (the target) has been the
center-stage in education research. Our concept of schooling is based
on the assumption that what children learn early on impacts their
learning and actions in the later grades.
The benefits
of knowledge transfer has engaged cognitive psychologists for over a
century. Although this underlying philosophy has also been a long-time
inspiration to machine learning and artificial intelligence in general,
there has been a renewed interest in knowledge transfer only recently.
It is now an active research area that requires new approaches,
formalisms, algorithms, and testbeds. Traditional machine learning
considers problems drawn from the same “population”, while transfer
learning considers problems from different populations and thus raises
machine learning to a higher and more challenging level. Transfer can
be especially effective when the learned knowledge is suitably
structured, for example, in a relational or hierarchical fashion. This
workshop is devoted to transfer learning in all subareas of machine
learning, including but not limited to, concept learning, clustering,
reinforcement learning, analogy, and to applications and evaluation
methodologies of transfer learning, with an emphasis on how the learned
knowledge is structured and exploited.
